[ExI] Watson On Jeopardy.
rpwl at lightlink.com
Thu Feb 17 18:50:17 UTC 2011
Eugen Leitl wrote:
> On Wed, Feb 16, 2011 at 08:13:59PM -0500, Richard Loosemore wrote:
>> Distributed Processing", and if you have then read my papers, and if you
>> are still so much in the dark that the only thing you can say is "I
>> haven't seen anything in your papers that rise to the level of computer
>> science" then, well...
> You know, I could rattle off a list of books (far more relevant)
> you have no clue of. It's a pretty stupid game, so let's not play it.
If you actually read the thread you will see that nobody was playing
that "pretty stupid game", before you started to do so in the above
You have drastically, utterly failed to understand or read the context.
As I will explain....
I was addressing an implicit question from Kelly Anderson about how
anyone could make sense of my *own* papers, and I pointed to those two
books because I am claiming that they represent a critical hinge point
in the history of cognitive science and AI, and my work is best
understood as a path-not-taken from that hinge point.
If you think you know my own research better than I do, and can "rattle
off a list of far more relevant books" that would help someone
understand the context that my work comes from, by all means do so.
Granted, there is a problem there. Quite a few computer science people
read those McClelland and Rumelhart books looking only at the NN
algorithms, but without knowing the cognitive psychology history that
came before the PDP books. The problem is that my work springs not from
the superficial NN stuff but from that much deeper history. That fact
may cause some misunderstanding.
In order to gauge the appropriate level at which to respond to Kelly's
concerns, therefore, it mattered a good deal whether he was a cognitive
psychologist or an AI person, and to that end I went on to explain that
and ask some questions.....
>> At this stage, what you can get is a general picture of the background
>> theory. That is readily obtainable if you have a good knowledge of (a)
>> computer science, (b) cognitive psychology and (c) complex systems. It
> I don't see how cognitive psychology is relevant. It's good that
> complex systems makes your list.
Again, I mentioned cognitive psychology only because I was responding to
Kelly's comment about the fact that he read my papers but could not see
in them the things I had hoped he would. I was in the process of
explaining the background to my own work.
You seem to have interpreted my reference to those areas as something
Cognitive psychology is critical to an understanding of my approach to
AGI. Without an understanding of that field, it might be hard to see
why the papers I wrote are outlining a theory of AGI.
>> also helps, as I say, to be familiar with what was going on in those PDP
>> Do you have a fairly detailed knowledge of all three of these areas?
> Are you always an arrogant blowhard, Richard?
Do you always make comments like these without having read the messages
that came just before the one you are responding to?
To repeat, I was asking the question of Kelly because it was directly
relevant to his own comments about my papers. I needed to get a context.
Kelly responded politely and factually. You, on the other hand, are an
onlooker who the question was not directed at, but you feel inclined to
step in, misinterpret the context, and start using comments like
(... the kind of language that, I might point out, has been used as
grounds for putting people on moderation! ;-) ).
>> Do you understand where McClelland and Rumelhart were coming from when
>> they talked about the relaxation of weak constraints, and about how a
>> lot of cognition seemed to make more sense when couched in those terms?
>> Do you also follow the line of reasoning that interprets M & R's
>> subsequent pursuit of non-complex models as a mistake? And the
>> implication that there is a class of systems that are as yet unexplored,
>> doing what they did but using a complex approach?
>> Put all these pieces together and we have the basis for a dialog.
>> But ... demanding a finished AGI as an essential precondition for
>> behaving in a mature way toward the work I have already published...? I
>> don't think so. :-)
> I think two things apply: you haven't build a lot of systems that
> make impressive results, and you spend a lot of time on this list,
> which means you don't have have a lot of quality time for work,
> whatever it is.
> I've just skimmed your papers at maximum speed, and preliminary impression
> is not good. I'll reserve my opinion until I can read them.
Sadly, I can tell you in advance that your opinion will be of no value. :-(
More information about the extropy-chat